Ideas to speedup SweepFormula selrac, selsci

[MichaelHuth]

First #2154 should be resolved.

• For RAC currently over all initial selections is iterated, but it only needs to iterate over all unique sweep numbers per experiment/datafolder/device.

• RAC and SCI result in new additional selections. These new selections can contain selections that are identical to initial selections. If that is the case then we know that these selections are already of the same RAC or SCI. Thus, we could skip AFH_GetSweepsFromSameSCI / AFH_GetSweepsFromSameRACycle for initial selections that we already got as a result of a previous request.